[ONGOING WORK]

Authors

Roland Bouillot - Université Catholique de Louvain & Maastricht University 

Bertrand Candelon - Université Catholique de Louvain 

Clemens Kool - Maastricht University 

Abstract

Financial fragmentation in the Euro Area has become a hot topic since the European Central Bank must decide whether to maintain its interest rates high and long enough to tame inflation or loosen its monetary policy to mitigate the risk of another European sovereign debt crisis. This study leverages a novel high-dimensional dataset covering a wide range of fields within 10 European countries over the 2007 to 2024 period. This Big Data dataset has been exploited by a new machine learning technique (XGBoost) to find evidence of the financial fragmentation risk in the Euro Area. Most importantly, our main result suggests that the predicted long-term yield spreads of peripheral countries rise while those in core countries’ rise remain contained or even decrease. This divergence in yield differentials put a lot of scrutiny on the action of the ECB and call for policy guidance to avoid a new European sovereign debt crisis.

Keywords: Machine Learning, Financial Fragmentation, XGBoost, Sovereign spreads